Applications Of RVE, AQE Models In SINS Alignment On Moving Base

One of the key technologies of SINS is the initial alignment, whose accuracy affects the technique performances of navigation of the strap-down inertial navigation system (SINS). Generally, Selfalignment method is adopted in the initial alignment of SINS, that is, Selfalignment is realised by using accelerometer and gyroscope measuring the acceleration of gravity and the angle speed of the earth's rotation, but there exist many kinds of stochastic factors in the system, the precision and the interval of the initial alignment conflict, the higher precision required, the longer time needed.The rotation vector error model is derived based on quaternion attitude error model, and the relationships are shown among rotation vector, quaternion and direction cosine matrix. The nonlinear alignment model was built according to the rotation vector error model and linear velocity error model. Unscented Kalman Filter is applied in the simulation of alignment. The concept of the unification of rotation vector is brought up, and solved with the unification of quaternion. Finally the simulation results show that the model error is predominant in linear error model based alignment. But, in nonlinear error model based alignment, sensors bias and noises are predominant. The advantage of rotation vector error model based alignment is shown in following three aspects: static accuracy, initial angle of deviation and convergency time in comparison with linear error model based alignment.The initial alignment of vehicle in moving state is studied mostly in this paper aiming at fast response of the vehicle carried SINS. Since the accuracy and the start-up interval of INS have much relation to the initial alignment, the technique of the initial alignment of INS has become one of the research hotspots in the filed of inertial navigation recently. This paper studies the initial alignment algorithem of SINS.The main work is as follows:Firstly, multi-models for the alignment of SINS in the moving state are established: AQE, RVE,Ψangle attitude error model. Secondly, the H_∞robust filter and Unscented Kalman Filter (UKF) are established.Thirdly, the observability of SINS is analyzed .In the observability analysis, the model is considered as a constant system in the stationary base alignment and the theory of piece-wise linear constant system is introduced for alignment of SINS in moving state.Lastly, apply theΨangle attitude error model and the RVE in the In Flight Alignment (IFA), the results indicate that the RVE have better overall profromence than traditional attitude model.